/LKPNR

Primary LanguageJupyter Notebook

1. data sampling

cd NNR

bash download_extract_MIND.sh.

2. Data process

2.1. The Embedding by LLM

./NNR/pretrain_emb/item_emb.npy

2.2. Save the adjacent entity dictionary, leaving only adjacent entities with emb vectors

./graph/count_link_count.ipynb

2.3 Saving entities appearing in news&embedding entities in merged datasets

./graph/get_node_emb.py
./graph/get_node_emb.ipynb

3. model training

python main.py --news_encoder=MHSA --user_encoder=MHSA

4. Ohters

The necessary files for running the program can be downloaded from https://pan.baidu.com/s/13AVXSWxIuXI14UgKjaSpbg?pwd=siuf